Loan credit line determination method, electronic device and storage medium

Abstract

The invention relates to a loan credit line determination method, electronic equipment and a storage medium thereof. The method comprises the following steps: S1, acquiring original data related to a user to be credit-granted; s2, extracting a plurality of derivative features from the original data through feature learning, and establishing a user portrait system through the plurality of derivative features; s3, data preprocessing is conducted on the original data and the multiple derivative features, the original data and the derivative features are divided into a training set and a verification set, and the proportion of the number of data in the training set to the number of data in the verification set is 7: 3; s4, constructing an XGBoost tree integrated prediction model; s5, optimizing hyper-parameters of the XGBoost tree integrated prediction model; s6, determining the credit line interval of the target customer, determining the final weight of the customer, and giving the final line of the customer; step S7, carrying out credit line prediction on the verified user; and step S8, marketing is carried out on the credit-granted clients without dynamic expenditure, and dynamic expenditure of the credit-granted clients is promoted.

Description

technical field [0001] The present invention relates to the technical field of Internet finance, in particular to a method for determining a loan credit line, electronic equipment and a storage medium thereof. Background technique [0002] In today's financial industry, before financial platforms provide customers with credit services, they must first formulate credit policies that match customers. At the same time, the generation of credit policies is the key to risk control and revenue growth in the credit life cycle. In the method of the prior art, there is an unreasonable credit strategy caused by inaccurate data, which affects the accuracy of the risk assessment results and increases the risk of credit granting. The existing online or offline credit decision-making methods still have problems such as single indicators, untimely update of decision-making models, and model mismatches, which lead to increased risks for credit based on credit decision-making. [0003] At t...

AI Technical Summary

Problems solved by technology

[0007] 2. The existing online or offline credit decision-making methods still have problems such as single indicators, untimely update of decision-making models, model mismatch, etc., resulting in increased risks for credit based on credit decision-making; and

Images